Simplifying Best Practices into BI Development€¦ · 19/02/2012 · Each best practice needs a...

Simplifying Best Practices into BI Development

By Gary Hanson, Principal Consultant, Analytic Data Solutions, LLC

February 20, 2012

Table of Contents

OVERVIEW .........................................................................................................................................1

DEFINITION OF A BEST PRACTICE ........................................................................................................1

COMMON BEST PRACTICES FOR BI SOLUTIONS ...................................................................................2

ESTABLISHING A FORMAL ‘BEST PRACTICE’ PROCESS ..........................................................................3

TAKING OWNERSHIP MAKING EACH BEST PRACTICE UNIQUE ..............................................................4

BENEFITS OF A BI CENTER OF EXCELLENCE ..........................................................................................5

CONCLUSION .....................................................................................................................................6


Copyright 2012, Analytic Data Solutions, LLC. All rights reserved. Page 1

AnalyticData Solutions

BI Methodology

Simplifying Best Practices into

BI Development

Top Performing Best Practice

For BI Solutions

An Essential Best Practice – a deeper

Examination of:

Data Architecture

Data Model1 2 3, 4, & 5

Overview

A goal of all development projects is to implement the solution on-time and on-budget. Unfortunately,

there are too many times that we miss that goal. A common reason for significant project delay is when

the project team utilizes either an incomplete or an inferior design technique. Needless to say, it takes

time and money to rework or reengineer a solution. The later in the development process a problem is

discovered, the more expensive it is. More often than not, following a proven best practice can help to

avoid costly mistakes, keeping the project on-schedule and on-budget.

A cause of project delays may come from the tendency to rush through design activities by either cutting

corners or not properly completing a technique. The ‘rushing’ may also be attributed to not planning

adequate time as well as not understanding the design technique well enough. Subsequently, a significant

benefit of best practices is that it contributes to a form of mentoring and educating of your staff to the

proper techniques, ensuring consistent results. Your investment of maintaining and evolving best

practices will pay dividends in delivering timely BI solutions.

To properly address the breadth of this theme, best practices, this paper is first in a series of white papers

taking you through a progression as depicted below:

The purpose of “Simplifying Best Practices into BI Development” is to highlight the significance of

establishing and incorporating best practices into both Data Warehouse (DW) and Business Intelligence

(BI) development. A key component is the discipline of capturing your experiences or lessons learned

back into the best practice itself. This enhances the value add proposition of best practices and

contributes to achieving the highest level of business capabilities via BI solutions.

Definition of a Best Practice

Since best practices are used in nearly every industry and professional discipline, it is important to make a

distinction that, for this paper, the term ‘best practice’ is used within the boundaries of both DW and BI

development typically found in an Information Technology (IT)

environment. Let’s start by using a common definition of a best

practice provided by Wikipedia – see right. To achieve the lofty goals

of producing superior results and improving or evolving over time, a

best practice requires a thorough and complete structure ensuring

breadth and completeness.

Each best practice needs a general description highlighting its purpose

followed by a set of justification and rationalization rules. It is also

important to describe the implications or consequences of not

performing this best practice. Finally, identify any known exceptions

to this best practice and give examples. By combining the purpose

and rationalization, along with its implications and exceptions, you

will bring better balance in how to use each best practice. In addition, via the process of capturing

lessons learned and enhancing the best practices, you will allow it to evolve and adapt into the

mainstream process.

As defined by Wikipedia:

A best practice is a method

or technique that has

consistently shown results

superior to those achieved

with other means, and that is

used as a benchmark. In

addition, a "best" practice

can evolve to become better

as improvements are

discovered.




Best practices play an important role in the creation and execution of a Business Intelligence Competency

Center (BICC) or BI Center of Excellence. A primary purpose of a BICC is to facilitate and enable the

delivery of high quality BI solutions (business capabilities) by ensuring that best practices are established,

shared, and applied throughout the organization.

A BICC can help to initiate and reinforce predictable, repeatable, and consistent use of Business

Intelligence functionality, in part, through the use of best practices. Centralized BI support functions and

services that utilize best practices result in a reduced total cost of ownership (TCO). Consequently, there

is a reduced risk of flawed project implementations and increased likelihood of successful projects.

Common Best Practices for BI Solutions

BI solution is a broad term typically used to describe some sort of business capability involving technology

and analytics. So before we get into the ‘how to incorporate best practices into your development

processes’, let’s talk a bit in terms of what is a BI solution and why create a best practice. I think it is safe

to say that most BI solutions have a fair amount of design complexity and business complexity associated

with them. When appropriate, these complexities will benefit the most by

applying repeatable and consistent best practices to them.

A way to determine if creating a best practice merits consideration is to

weigh its importance related to BI solutions as well as the level of risk

associated to the effort. Here are a few questions to ask:

Is it important to design certain components using a consistent and

repeatable technique? Is it important to follow a certain process using a

consistent and repeatable method? Do you have a rather inexperienced

team? Are the implications or consequences of not conforming to best

practices considered high risk? Is BI solution development occurring

across several organizations and/or geographically separated groups?

Best practices are meant to augment your development process and not replace it, so look for areas to

compliment work activities and avoid creating conflicting instructions. The following outline

summarizes a set of common best practices that I have used and found very valuable in developing DW

and BI solutions:

Your list of best practices may vary from the above as each organization will have different goals and

may place different emphases on the set of practices it considers most important. Subsequently, a key

take-away is that each best practice is branded to the way your organization performs it and that over

time, it is enhanced by applying your experiences to each one.

1. Establish a data design driven by business requirements and fostered by business stewardship

2. Adopt an iterative, flexible development methodology addressing DW and BI solutions

3. Adhere to the design rules of dimensional modeling for analytic data models

4. Allow only data that has business analytic value to be stored in the analytic data mart layer

5. Utilize a three tier data architecture to manage Enterprise data

6. Acquire data from the true data source or system of record

7. Leverage the corporate investment and functionality of strategic tools

8. Apply appropriate data reconciliation and data quality considerations to the data architecture

9. Apply appropriate access control and security considerations to the data architecture

10. Keep performance in mind when designing solutions rather than address it as an afterthought

BI solutions represent many

types of deliverables that

come from either DW or

Analytic Reporting

development projects. As

examples, a BI solution

could range from creating a

new data mart to delivering

a new analytic reporting

capability to deploying a

new metadata repository.




Establishing a Formal ‘Best Practice’ Process

It is important to note that establishing a formal ‘best practice’ process does not replace the need for a

strong development methodology. As the graph below illustrates, most organizations will have governing

bodies (i.e., Project Management Office & Enterprise Architecture) overseeing a development

methodology. As a complementary

process, best practices simply augment

the appropriate methodology. Each best

practice completes, in some fashion, the

overall development process. This way

everyone follows a consistent pattern and

thoughtful process.

A best practice is used to describe and

enhance a process or technique by

following a consistent and repeatable

method of doing things that multiple

organizations can take advantage of.

Unfortunately, it is often skipped by

companies, even though they may

consistently employ and follow a strong

development methodology. So it is

important to integrate each best practice into the appropriate methodology to ensure that it is used when

needed.

As stated in the definition, a best practice is a method or technique that has consistently shown results

superior to those achieved with other means, and that is used as a benchmark. As shown in the below

diagram, best practices cover or represent different aspects or features within the BI environment. It may

address specific design techniques, to broad architecture considerations, to data governance (business)

processes or different methods ensuring expected results:

A superior BI foundation is built on a combination of best practices and guiding principles that provide a

“foundation” for a long lasting Enterprise Analytics Environment.

Work with your PMO or project delivery team to incorporate formal sign-offs and approval points of

critical best practices within the methodology and where appropriate, utilize checklists to ensure that the

Project Management Office (PMO)

Enterprise Architecture (EA)

Development Methodology

Best Practices

SDLC DWLC

• Standards• SOA/Framework• Governance

• Process • Controls• Resources

Best Practices

Iterative BI Methodology

Data Governance

Metadata Management

Business Intelligence

Data WarehouseData Integration




required technique was performed and compliance was met. During the creation of the project plan,

make sure that you put in adequate time to each activity covering the best practice.

This development time-line illustrates the necessary touch points over the life-span of a development

project:

Documenting and maintaining best practices can be a complicated and laborious process, but will help to

avoid costly mistakes keeping projects on schedule and on budget. Your investment of maintaining and

evolving best practices will pay dividends in delivering timely BI solutions.

Taking Ownership making each Best Practice Unique

Taking ownership is a key ingredient to achieving and preserving the benefits of best practices. As the

purpose behind best practices are to trigger the use of repeatable and proven patterns of either design

techniques or process methods, it is important to find ways to improve and grow them. This enhances the

value add proposition of best practices and contributes to achieving the highest level of business

capabilities within BI solutions.

This ownership includes the discipline of capturing

your experiences or lessons learned back into the best

practice itself. At the end of each project, conduct a

formal exercise of reviewing the project successes and

issues with the full project team. Make sure you

structure these reviews as non-threatening and free of

any retaliation as your goal is to gain any and all

feedback. Only then can you start formulating the real

experiences and issues capturing lessons learned back

into the best practices.

Don’t think that you’ll simply have time to do this

exercise when the project completes as this rarely

happens. You need to anticipate the amount of time

needed and incorporate it in the project plan itself from the very beginning. Managing all aspects of

project development requires a lot of work including managing and growing best practices.

Don’t allow this to be optional as then it will never happen and make sure you include all team members

because everyone can and should contribute to this important feedback process. Another trick to ensure

completion of lessons learned is to incorporate this as a criteria item within the individual performance

review. This way ‘gaining ownership’ becomes part of the norm and not the exception to the day-to-day

work.

Pre-Req’ts Analysis Design

1) Be sure to provide adequate time to the ‘best practice’ activities within the project plan2) Employ aspects of an iterative methodology where appropriate – e.g., prototype or proof-of-concept

Code Test Implement

Timely execution, approvals & sign-offs

Improved Best

Practice

Current Best Practice

Lessons Learned




Benefits of a BI Center of Excellence

BI Center of Excellence or Business Intelligence Competency Center (BICC) is a significant theme in

itself and merits its own series of white papers. So here I’ll highlight the general benefits of a BICC as a

formal platform for best practices.

A BICC enables the leveraging of Business Intelligence across the organization. This is achieved through

strengthened BI tool skills, a broader and increased usage of BI, and the elimination or minimizing of re-

training. This diagram depicts the BICC service and product offerings via the means of consulting,

guiding, researching, and training:

Organizations with a BICC experience greater business user satisfaction. Contributing to the higher

satisfaction is an increase in self-service capabilities and the ability to react more quickly to changes in

the business. Having shared business knowledge and experience readily available, there is a greater

understanding and more accurate use of the data. More timely access to trusted information and an

increase in quality of data available for analysis, improves an organization's decision making ability.

A BICC can help to initiate and reinforce predictable, repeatable, and consistent use of BI. Rework is

minimized, being done correctly once and then shared with others. BI can enhance productivity. With

increased processing capabilities, more can be produced with existing staff. Projects consume fewer

resources with a shortened implementation time and increased throughput.

Centralized BI functions and services that utilize best practices result in a reduced total cost of ownership

(TCO). There is a reduced risk of flawed project implementation and increased likelihood of successful

projects. Use of technology resources and assets are maximized. By utilizing the current tools and

environment, investments and one-offs are minimized. A BICC promotes a standard, reusable BI

approach and architecture. Project and technical infrastructure costs are reduced, while leverage with

vendors is increased. By exploiting existing technological investments, software costs are decreased and

performance (run time) is improved. The cost of the overall BI environment is reduced.

ENTERPRISE ARCHITECTURE

PROJECT MANAGEMENT

APPLICATION DEVELOPMENT

SALES & MARKETING

ACTUARIAL & UNDERWRITING

BICC

ClassroomCBT

Designer: Developer:

Other

Training

IT Business

FINANCE

GOVERNMENT PROGRAMS INFORMATICS




Conclusion

The purpose of “Simplifying Best Practices into BI Development” is to highlight the significance of

establishing and incorporating best practices into both DW and BI development. It is important to

integrate each best practice into the appropriate methodology, ensuring that everyone follows a consistent

and thoughtful process. This enhances the value add proposition of best practices that contributes to

achieving the highest level of implementing business capabilities. Subsequently, your investment of

maintaining and improving best practices will pay dividends in delivering timely BI solutions.

Accepting ownership is a key ingredient to achieving and preserving the benefits of best practices. As the

purpose behind best practices is to trigger the use of repeatable and proven patterns of either design

techniques or process methods, it is important to find ways to improve and grow them. This ownership

includes the discipline of capturing your experiences or lessons learned after each project back into the

appropriate best practice, allowing improvements to be made on an on-going basis.

Also, a BICC or BI Center of Excellence can help to initiate and reinforce predictable, repeatable, and

consistent use of business intelligence functionality, in part, through the use of best practices. Centralized

BI functions and services that utilize best practices result in a reduced total cost of ownership (TCO).

Consequently, there is a reduced risk of flawed project implementations and increased likelihood of

successful BI solutions.

About the Author:

About Analytic Data Solutions, LLC:

Gary Hanson is the Owner and a Principal Consultant at Analytic Data Solutions, LLC. With over 30 years

in the Information Technology field and 15 years of practicing in the data warehouse and business

intelligence field, Gary has extensive hands on experience in information architecture, data integration, data

management, data modeling, and business intelligence. Gary is the co-author of Quick Start Analytics for

Healthcare and holds a B.A. degree in business management from Metropolitan State University

(Minneapolis, Minnesota).

Analytic Data Solutions is a BI-data solutions consulting company with a strong business orientation

specializing in creating a comprehensive enterprise analytic environment strategic roadmap or program plan.

This roadmap, coupled with the foundational components of the business requirements and the architecture

blueprint, provides the means to execute the appropriate sequence of projects achieving timely and business

value-add analytic solutions.

To learn more, please visit us at: http://www.analyticdatasolutions.net

http://www.analyticdatasolutions.net/

Simplifying Best Practices into BI Development€¦ · 19/02/2012 · Each best practice needs a...

Documents

Transcript of Simplifying Best Practices into BI Development€¦ · 19/02/2012 · Each best practice needs a...